RSS Feed Print
Solids Duplicate Determinations
CJ Felice
Posted: Monday, October 15, 2012 2:18 PM
Joined: 10/11/2012
Posts: 2


I'm new here and attempted to post this question last week, but it didn't display for some reason.

 

I have a question regarding Standard Methods 2540 Solids section.  Specifically, TSS 2540D, end of section 3.c.:  "Duplicate determinations should agree within 5% of their average weight."  Can someone please clarify this statement?

 

Does this mean to take the average of the two final weights and then compare each weight to the average?  If that comparison is within 5% of the average, then the values are acceptable?

 

How are you applying this in your benchsheet?  We have an Excel spreadsheet in which our criteria is set via formulas.  I would like to understand that statement so I can make sure it's already included in our sheet criteria. If not, I'll need to figure out how to enter it.

 

Thanks!


01605042
Posted: Wednesday, October 17, 2012 6:55 AM
Joined: 9/28/2009
Posts: 18


We record relative percent difference of duplicates on the benchsheet. RPD is the difference between the two samples divided by the average.

 

Excel formula  =ABS(A1-B1)/AVERAGE(A1,B1)
 


01487708
Posted: Wednesday, October 24, 2012 8:57 AM
Joined: 10/12/2009
Posts: 11


You are correct in that it is a comparison between the two residue weights and their average.  The Excel forumla would be

 

100*ABS(Cell1-Cell2)/(AVERAGE(Cell1:Cell2)*2)

 

Cell1 and Cell2 are the cells with the residue weights (final - tare)

100 is to covert to percent and 2 is because the comparison is to each weight.

 

Some may recognize the formula without the 2 factor as the one to compute % difference

 

This suffers the same problem as all others where the determination is against a fixed value in that there are ranges of weights where the comparison does not make sense.  Below are some actual results and how they fared

 

WT 1   WT 2   % differece to average

16.7   16.8          0.3

  2.2     3.2        18.5

  3.1     3.3          3.1

  6.5     6.9          3.0

 

As you can see the second pair compute to an 18.5% difference.  The sample is of a final effluent and 1000mL was sampled.  The only way to improve this would have been to use more sample and collect more residue.

 

The good news is the wiggle word "should".  "Should" does not mean "Shall" so you can write up a work around.  In these cases, I suggest the "If..then" process where the 5% applies unless some limit is met then another criteria is used.  For example

 

5% for weights over 5 mg, 1.0 mg for 5 mg or less